The currently prescribed insulin sensitizing drugs, (the full agonists rosiglitazone and pioglitazone), bind to peroxisome proliferators-activated receptor PPAR, a transcription factor) and improve maintenance of normal blood sugar levels, which helps prevent the long-term complications of diabetes, including heart disease and stroke for many of the 25 million diabetic Americans. The anti-diabetic effects of PPAR binding drugs are more durable (prevent diabetes progression for a longer time) and more effective than other drugs such as metformin and glyburide making PPA interacting drugs a valuable treatment for diabetes. However use of these drugs is limited because they cause increased bone fracture in women, weight gain, heart failure and edema. Effective PPAR targeted anti-diabetics with less unwanted effects are needed. PPAR is a transcription factor that interacts with many other proteins to carry out its effects, it interacts with these other proteins on its surface. Many of these interactions depend on a particular ligand being present in the large and hydrophobic PPAR ligand binding pocket. Somehow, binding of ligand changes the structure and/or dynamics of PPAR's interacting surfaces to favor or disfavor interaction with these partners (i.e. co regulators and heterodimerization partner, retinoid X receptor, RXR). This unique ligand induced (or ligand free) PPAR/partner complex increases or decreases the expression of many genes which leads to ligand specific in vivo effects. PPAR crystal structures bound to many different kinds of ligands, both alone and mated with RXR, co regulator proteins and DNA are very similar and do not explain the functional differences observed for different classes of PPAR ligands. We hypothesize that adding movement information to these crystal structures will help correlate PPAR regions and movements with in vivo effects of the distinct classes of PPAR ligands. This correlation map could then be used to drive more effective PPAR drug development. We have evidence that different ligand classes (full, partial and non agonists) can indeed be differentiated by the dynamics they induce. We have found that a PPAR binding non-agonist and two partial agonists all bind PPAR in at least two distinct orientations, while the ful agonist does not. Furthermore our PPAR movement data suggest a novel mechanism for partial agonism in PPAR. Connection of PPAR structure and movement to functional effects will help improve design of novel PPAR drugs with better separation of unwanted effects (e.g. heart failure) from wanted effects (i.e. anti-diabetic efficacy, anti-inflammation). This will improve diabetes treatment and open up possibilities for PPAR targeted therapies for atherosclerosis and autoimmune disorders.